The use of artificial intelligence to assess diabetic eye disease among the Greenlandic population
Background: Retina fundus images conducted in Greenland are telemedically assessed for diabetic retinopathy by ophthalmological nurses in Denmark. Applying an AI grading solution, in a Greenlandic setting, could potentially improve the efficiency and cost-effectiveness of DR screening.Method: We dev...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2024-12-01
|
| Series: | International Journal of Circumpolar Health |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/22423982.2024.2314802 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850220934836781056 |
|---|---|
| author | Trine Jul Larsen Maria Bråthen Pettersen Helena Nygaard Jensen Michael Lynge Pedersen Henrik Lund-Andersen Marit Eika Jørgensen Stine Byberg |
| author_facet | Trine Jul Larsen Maria Bråthen Pettersen Helena Nygaard Jensen Michael Lynge Pedersen Henrik Lund-Andersen Marit Eika Jørgensen Stine Byberg |
| author_sort | Trine Jul Larsen |
| collection | DOAJ |
| description | Background: Retina fundus images conducted in Greenland are telemedically assessed for diabetic retinopathy by ophthalmological nurses in Denmark. Applying an AI grading solution, in a Greenlandic setting, could potentially improve the efficiency and cost-effectiveness of DR screening.Method: We developed an AI model using retina fundus photos, performed on persons registered with diabetes in Greenland and Denmark, using Optos® ultra wide-field scanning laser ophthalmoscope, graded according to ICDR.Using the ResNet50 network we compared the model’s ability to distinguish between different images of ICDR severity levels in a confusion matrix.Results: Comparing images with ICDR level 0 to images of ICDR level 4 resulted in an accuracy of 0.9655, AUC of 0.9905, sensitivity and specificity of 96.6%.Comparing ICDR levels 0,1,2 with ICDR levels 3,4, we achieved a performance with an accuracy of 0.8077, an AUC of 0.8728, a sensitivity of 84.6% and a specificity of 78.8%. For the other comparisons, we achieved a modest performance.Conclusion: We developed an AI model using Greenlandic data, to automatically detect DR on Optos retina fundus images. The sensitivity and specificity were too low for our model to be applied directly in a clinical setting, thus optimising the model should be prioritised. |
| format | Article |
| id | doaj-art-b2d456a6b50a4addbc9928c6eced1bf5 |
| institution | OA Journals |
| issn | 2242-3982 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | International Journal of Circumpolar Health |
| spelling | doaj-art-b2d456a6b50a4addbc9928c6eced1bf52025-08-20T02:06:51ZengTaylor & Francis GroupInternational Journal of Circumpolar Health2242-39822024-12-0183110.1080/22423982.2024.2314802The use of artificial intelligence to assess diabetic eye disease among the Greenlandic populationTrine Jul Larsen0Maria Bråthen Pettersen1Helena Nygaard Jensen2Michael Lynge Pedersen3Henrik Lund-Andersen4Marit Eika Jørgensen5Stine Byberg6Greenland Center of Health Research, Institute of Nursing and Health Science, University of Greenland, Nuuk, GreenlandClinical Epidemiology, Steno Diabetes Center Copenhagen, Copenhagen, DenmarkClinical Epidemiology, Steno Diabetes Center Copenhagen, Copenhagen, DenmarkGreenland Center of Health Research, Institute of Nursing and Health Science, University of Greenland, Nuuk, GreenlandClinical Epidemiology, Steno Diabetes Center Copenhagen, Copenhagen, DenmarkSteno Diabetes Center Greenland, Nuuk, GreenlandClinical Epidemiology, Steno Diabetes Center Copenhagen, Copenhagen, DenmarkBackground: Retina fundus images conducted in Greenland are telemedically assessed for diabetic retinopathy by ophthalmological nurses in Denmark. Applying an AI grading solution, in a Greenlandic setting, could potentially improve the efficiency and cost-effectiveness of DR screening.Method: We developed an AI model using retina fundus photos, performed on persons registered with diabetes in Greenland and Denmark, using Optos® ultra wide-field scanning laser ophthalmoscope, graded according to ICDR.Using the ResNet50 network we compared the model’s ability to distinguish between different images of ICDR severity levels in a confusion matrix.Results: Comparing images with ICDR level 0 to images of ICDR level 4 resulted in an accuracy of 0.9655, AUC of 0.9905, sensitivity and specificity of 96.6%.Comparing ICDR levels 0,1,2 with ICDR levels 3,4, we achieved a performance with an accuracy of 0.8077, an AUC of 0.8728, a sensitivity of 84.6% and a specificity of 78.8%. For the other comparisons, we achieved a modest performance.Conclusion: We developed an AI model using Greenlandic data, to automatically detect DR on Optos retina fundus images. The sensitivity and specificity were too low for our model to be applied directly in a clinical setting, thus optimising the model should be prioritised.https://www.tandfonline.com/doi/10.1080/22423982.2024.2314802Diabetic retinopathyartificial intelligencescreeningultra wide-fieldICDR-scale |
| spellingShingle | Trine Jul Larsen Maria Bråthen Pettersen Helena Nygaard Jensen Michael Lynge Pedersen Henrik Lund-Andersen Marit Eika Jørgensen Stine Byberg The use of artificial intelligence to assess diabetic eye disease among the Greenlandic population International Journal of Circumpolar Health Diabetic retinopathy artificial intelligence screening ultra wide-field ICDR-scale |
| title | The use of artificial intelligence to assess diabetic eye disease among the Greenlandic population |
| title_full | The use of artificial intelligence to assess diabetic eye disease among the Greenlandic population |
| title_fullStr | The use of artificial intelligence to assess diabetic eye disease among the Greenlandic population |
| title_full_unstemmed | The use of artificial intelligence to assess diabetic eye disease among the Greenlandic population |
| title_short | The use of artificial intelligence to assess diabetic eye disease among the Greenlandic population |
| title_sort | use of artificial intelligence to assess diabetic eye disease among the greenlandic population |
| topic | Diabetic retinopathy artificial intelligence screening ultra wide-field ICDR-scale |
| url | https://www.tandfonline.com/doi/10.1080/22423982.2024.2314802 |
| work_keys_str_mv | AT trinejullarsen theuseofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT mariabrathenpettersen theuseofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT helenanygaardjensen theuseofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT michaellyngepedersen theuseofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT henriklundandersen theuseofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT mariteikajørgensen theuseofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT stinebyberg theuseofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT trinejullarsen useofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT mariabrathenpettersen useofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT helenanygaardjensen useofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT michaellyngepedersen useofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT henriklundandersen useofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT mariteikajørgensen useofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation AT stinebyberg useofartificialintelligencetoassessdiabeticeyediseaseamongthegreenlandicpopulation |